This was written as a response on lenr-forum.com, but became long and I decided this is worth a post here.
AlainCo wrote: If you came with 1% error, and an anomaly of 10%, a skeptic will bash you as “you have unaccounted error sources”, which is probably right, but not by 10%.
I consider it necessary in this field to stop worrying about skeptics, be skeptical oneself, and design controlled experiments for testing hypotheses, including exploring the parameter space, not for proving things to skeptics.
This is why calibration, pre and post, and even during, are required.
Yes. Calibration is a variety of controlled experiment. Calibration measures the behavior of the equipment. As well, in a relative frenzy to demonstrate More Heat for skeptics, it was lost that correlation can punch through noise so high that the signal from a specific experiment can’t be visually recognized. This requires many measurements, under controlled conditions. It is the opposite of quick and dirty.
To some extent, we viewed skeptics as the Enemy, and when that thought is active, our intelligence is dominated by survival responses. It’s primitive thinking and generates primitive results.
I remember experiments done by F&P, probably McKubre Storms, where the calorimeter is controlled as stable during the experiments with a heat pulse.
F&P, in particular. Some back-story: Pons and Fleischmann were exploring the possibility of a deviation in fusion rate from prediction through 2-body quantum mechanics, due to the much more complex conditions of condensed matter. They thought that such a deviation would be small, possibly beyond their ability to measure it, so they designed very precise calorimetry. The impression that I have, apparently shared with Miles and McKubre, is that the F&P calorimetry was the best in the world, and the precision or accuracy is said to be plus or minus 100 microwatts.
This path was not followed, generally, by other workers, perhaps because F&P calorimetry was computationally complex. Their calorimetry was called isoperibolic, from the constant-temperature bath they immersed their cells in. Strict isoperibolic calorimetry requires stable conditions, for temperature to settle, so that there is constant heat flow between the material being tested and the mass at constant temperature. That takes substantial time, and in these experiments the generated heat may be quite noisy (as well as input power, during electrolysis, because of bubble noise and other variations).
So their calculations included a heat capacity term, where the rate of change of temperature is used to infer power. That is normally called adiabatic calorimetry. That is, their calorimetry was hybrid. They used a half-silvered Dewar flask for their cell, so that heat flow was through a part of the cell that was not silvered, that was always fully immersed in the heavy water, and also in the constant-temperature bath.
There are two issues to consider: precision of calorimetry and the communication of results. Papers often confuse the two. That is, in a paper reporting excess heat — or lack of same — there may be substantial detail on how the heat was measured, with perhaps complex formulae, when the same complications are not reported for measuring, say, voltage. Yes, the “reason” for this is obvious. They are not using an Acme Microwatt Precision Calorimeter, with known precision and calibrated accuracy. However … it would be entirely possible to study the calorimeter and calorimetry itself, entirely aside from any question about excess heat, measuring not only calibration pulses (i.e, F&P would regularly inject a known power through a resistor, and observe the effect on their calculated power), but known chemical heat. Calibrate the hell out of it, and use many points of temperature measurement, so that a shift in position of heat generation can be tested. Thoroughly study it, and publish that (anywhere; however would probably best be an electrochemmistry journal or the like, i.e., a journal that is on-point for what is studied.
In an excess heat paper, simply refer to that as the method of calorimetry, and report the Labview XP results. Make the raw data accessible in supplemental material. Show single-variable correlations. How did XP vary during the experiment? How did it correlate with temperature, and with other variables, such as the resistance behavior of the cathode (end to end, correlated with loading ratio)? And if you mention COP, duck, because I’ll be tossing rotten tomatoes.
COP is a measure that is inordinately confusing, it exists only to attempt to convince skeptics. McKubre used 5% of input power as a standard below which his results were not considered significant. That is partly because he was using a conceptually simple but precision-clunky method of calorimetry, with precision more than two orders of magnitude worse than the F&P reputation.
For exactly the same underlying effect, COP can vary from low to very high, even infinite. Imagine a reaction generating a few watts, for a time, inside a bomb calorimeter. The bomb calorimeter has an initial interior temperature. Trigger the reaction. How? What if the reaction itself requires some temperature? The initial temperature of the reaction is an environmental condition. If a lab is studying the heat of some reaction, they do not consider the input power to their lab heating system as input power to the reaction. If a bomb calorimeter is inside an environmental chamber, the heating of that chamber is not relevant to reaction calorimetry, as long as the calorimetry is entirely interior to the bomb (and the heating is properly controlled).
Pons and Fleischmann could not use a bomb calorimeter and pure adiabatic calorimetry, because the experiment, from electrolytic power, would become entirely too hot, quickly. They needed something that would work for long electrolytic experiments, so there must be a cooling path.
However, consider the current Parkhomov cell. He heats the reactor tube with a coil wrapped around it. His interior temperature is measured with a thermocouple that is inside the cell, off to one side. It is not possible to clearly distinguish, in this experiment, the heating power from excess heat generated in the fuel, the heating is intimate with respect to a fuel effect. He also controls the heating thermostatically, and as a result, if one looks at his raw data rather than what he presents, which is some kind of average, it’s a huge mess. If his control is more linear, based on rate of change of temperature, this could be better, much better, but I’ve seen no sign that he is doing that. A redesign would separate and distinguish between heat from the coils and heat from the cell.
Songsheng Jiang attempted that, the big problem being poor choice in thermocouple type, he was operating outside of design conditions, so they failed. (And then he attempted to interpret results from failed or failing thermocouples.) But the basic concept was much better. (He also, however, did not control his cooling conditions. In the middle of the experiment, he turned on a cooling fan, because the entire apparatus was getting too hot. Is it a coincidence that this was the point where he started to see what looked like excess heat? He also changed pressure at the same time. His work was exploratory only, not definitive. He was changing multiple conditions at the same time. If one wants definitive results, Bad Idea. For exploration, do whatever you like!
However, if heating power is only used to raise environmental temperature, this is not “input power” to the reaction. If at some temperature the fuel heats up, gets even hotter, this is generated power (chemical or otherwise). COP is infinite. What becomes interesting is excess energy, because high power is possible from chemical reactions — for a short time. Energy is integrated power. So, then, we want to see time behavior.
I think if I see one more Ragone plot, I’m going to get sick. Please, no more!. They add nothing more than a statement that the calculated energy release from a sample was X joules per gram, whereas the highest known chemical potential energy is Y joules per gram. Ragone plots are not a sober part of an experimental report. They are part of an argument with skeptics, polemic.
I remember reading the progress reports of Edmund storms where he reported not only the results but the stability of the calorimeter constant to shows Shanahan style of calorimetric constant shift did not happen.
Shanahan has poorly communicated his critique. His “CCS” is not an actual shift of calibration constant, the name confuses anyone experienced with calorimetry (and this is quite visible in response to him from such people), but is the result, he proposes, of a change in location in the cell where heat is generated, plus, he tosses in, in some cases, additional heat from unexpected recombination inside an open cell. He is aware that this is unexpected and “extraordinary.” He is actually confirming the AHE, merely suggesting a prosaic cause, an unknown chemical or physical anomaly. His proposals are testable.
(And the general cause of “change in location” in closed cells would be recombination at the cathode rather than in the recombiner. This recombination from combustion of deuterium at a palladium cathode that is immersed in electrolyte is an “anomaly,” quite unexpected. Recombination occurs when deuterium and oxygen come into contact with palladium that is not immersed. Miles talks about pieces of co-deposited palladium falling off the cathode and then floating around on the surface, with flame from recombination of the released deuterium.)
Storms did do some work where he had two calibrations, one through Joule heating in the electrolyte and one through a resistor in roughly the same space as his recombiner catalyst. He found very similar calibration constants; there was a difference, but small. Shanahan, if I’ve understood him correctly, ignored that.
what could be nice would be heat after death too.
And now we get to what is indeed really nice, and of high practical import.
Storms recently showed a new kind of HAD. This should have been done many years ago! He maintained the electrolyte temperature with a heater. He loaded his cathode normally, until he was seeing XP. Then he turned off the electrolytic current, but continued to maintain the electrolyte temperature.
(This took less heating than without XP, I’d assume, but the XP was not nearly enough to maintain that temperature by itself. To do that would take much more insulation, and, at this point, increasing the experimental complexity to create such is not worth the effort. Convincing skeptics is simply not a major motivation for this work. There is another way, much more powerful, to accomplish that. I will give that way below.)
He found that, even though loading declined with time, because removing the electrochemical “pressure” would cause release of deuterium, the XP was maintained.
This is, in any case, how I have interpreted his recent work, and I’d appreciate corrections. The work is detailed in his most recent paper in JCMNS. See page 96 of the PDF for his report of HAD.
How to convince skeptics? Very simple. Publish in Nature. I expect we may see that next year, or not long after. There is an obvious precondition: write a paper reporting conclusive, verifiable work, of high significance, including a review showing that this is merely a confirmation with increased precision of work that was first done long ago, and showing new experimental work by more than one research group. Nail it.
(If Nature refuses to publish such, i.e., continues an editorial policy about cold fusion established by 1990 or so, then something else happens, and the only difference is that Nature ends up with egg all over its corporate face. I don’t expect it. They will instead say, “We have been waiting for this!” as they proclaim that they were right to be skeptical until something like this was done. And I, for one, would be complimenting them for their courage, etc.)
3 thoughts on “COP is a red herring – what matters is reality of an effect”
If one cares about the result, it’s not science, but survival. Survival is obviously “not bad,” but … is also not science, which does best when we, at least, suspend caring and go for careful and detached observation. What the enterprise of science has shown that this reserve leads to better science and very likely stronger survival, if “not caring” doesn’t itself take over, as in having no regard for the effects of what we do.
There is one reason why someone with very solid result and energy to face Nature would not publish : to go commercial or at least to investigate more to propose a theory or bette result (thus going dark).
This domain is cursed by it’s disruptive nature.
Science is better done if nobody care of the result.